This is a paid article which your subscription is allowing you to read.

Stock markets have been soaring on strong company results and massive capital investment into AI. Some analysts, however, warn of a potential AI bubble, arguing that overinvestment in infrastructure may outpace revenue. On the other hand, the potential of AI to transform economies and markets leaves investors anxious not to miss out on the current boom. There may be a middle road that investors could consider that allows them both to buy into the boom while mitigating the risk of being exposed to individual company reversals.

AI market bears also flag that circular investment flows risk inflating valuations without immediate utility, reminiscent of the dot-com era, and warn that crashes could occur as overbuilt infrastructure struggles to deliver returns.

The bulls, however, argue the AI boom is real. Rapid adoption, rising token usage, and transformative applications suggest genuine demand. Current investments are laying the foundation for a long-term AI-powered economy, akin to the chaotic early investments during the industrial revolution that ultimately yielded lasting societal and economic shifts.

Amid this uncertainty, some market watchers suggest the more durable opportunity lies beneath the surface – in the systems that power AI.

Utilities, pipelines, rail networks, data centres, and telecommunications towers are the backbone of everyday life and now, increasingly, of AI itself. Their revenue streams are typically regulated, stable, and far less volatile than those of AI companies chasing growth targets and investor hype.

The surge in AI adoption is already driving energy demand. Data centre electricity consumption in the United States, under some projections, is expected to rise threefold by 2030 as hyperscale operators expand their computing infrastructure. Utilities and power providers stand to benefit directly, with long-term contracts offering predictable cash flow even if AI growth slows. Public and private infrastructure alike offer strong returns, with public assets providing liquidity and private holdings giving exposure to regulated sectors. Beyond AI, pipelines, railroads, and nuclear and renewable energy projects offer additional growth avenues.

For investors bullish on AI, the question is no longer just which AI stock to buy, but which foundational assets will sustain the AI boom. By focusing on the infrastructure underpinning this technology, portfolios can capture growth while cushioning volatility – a defensive yet forward-looking strategy in an era of both excitement and caution around AI.

Below those assets and utilities are even more foundational investment options into commodities such as copper, for example – demand for which may lift in line with expanding data centres and energy demand. Consider, for example, that because AI data centres will need such vast amounts of power, that will in turn require immense quantities of copper for power distribution, cabling, and cooling systems, but that copper supply is, to an extent, constrained due to aging mines and long, complex, and often difficult permitting processes for new projects. So influential may the AI boom be to copper that some analysts no longer view it solely as a cyclical industrial metal, but rather a critical strategic asset that may snap out of its historical price patterns – although constrained demand and high prices may also incentivise developing copper alternatives.

The upshot is very much that for investors looking to get into the AI boom, but concerned at the outlook and valuations of specific AI companies, there are alternatives that offer exposure to the phenomenon but at a different risk profile than buying into AI firms directly.